Thanks for taking a look. Please see the linked PDF drawings, and tell me if I have overlooked something important.

Existing Design (PDF)
Proposed Changes (PDF)

The existing design has two zones with excessively long 600' loops of pex hung between joists with plastic Wirsbo pex-clips. The tubing sits in a 3" airspace covered with only R13 batting (no foil). The major flaw I see is the continuous loops which should be no more than 300' long. Even at full boiler temperature, it is not putting out enough heat.

Proposed changes would use aluminum Joist-Trak on three 200' loops each, balanced with Mr Pex manifolds. R19 would fill the joist cavities. I plan to use one of the existing circulators as a primary pump, and feed two secondary loops from it, with 1" copper. The temperatue of each secondary loop would be controlled by hand-set thermostatic mixing valves. Using the ADS software I've determined the loops would call for about 97 degrees.

So, each thermostat will control its respective secondary loop circulator. But either one will also trigger the primary loop circulator and fire the boiler.

1) Is the location of primary loop circulator acceptable downstream from the secondary branches as shown?

2) Could this type of setup lead to short cycling the boiler?

3) To control the new setup, can I use a spare two-zone Taco control box? It would recieve the signals from the two thermostats, and switch on the secondary loop circulators. The relay output from that 2nd box (which is normally used to fire a boiler) would instead provide signals to the existing Taco box, controlling the primary pump and boiler. Am I missing something?

I understand that without motorized mixing valves that I might lose some efficency from the lack of automatic seasonal reset. Any advice is greatly appreciated.

 

I wouldn't run that primary loop through the boiler like that. The boiler becomes a point of heat loss when the burner shuts off and the pump is on.
Look into using a hydraulic separator and you'll simplify the whole operation.

 

Master, I'm not sure what you mean? If the burner cuts off on high limit and there is still a heat call, of course the pump will still be running... but when the heat call ends, the boiler AND pump will shut off. At least that's what I'm understanding from the description... are you seeing something I'm missing?

 

There ARE a couple fine points that I might do differently:

Maybe the drawing shouldn't be taken 'literally', but I would not mount the expansion tank facing up. Air will get trapped on the water side of the bladder and will cause the tank to rust from the inside out at the air/water interface point. Mount the tank HANGING from piping.

Perhaps that boiler is designed such that it will act as the 'air scoop', and the air will collect at the top of the boiler and be eliminated by a vent at that location... I don't know, haven't looked at that install manual lately... but what I don't see on this system is any other type of air removal device.

I probably would not pipe the cold water supply to the boiler on the return manifold. Nor would I place the circs pumping toward the boiler. I think I would put the pumps on the supply manifold, pumping away from the boiler, use pumps with IFC, get rid of the Taco flo-cheks that way... and place a proper air eliminator on the supply pipe out of the boiler. I would put the expansion tank at that location to create a PONPC, and also pipe the cold water supply in at that point.

Wherever you pipe the cold water supply, you want to install a valve between the boiler and the pressure reducing valve for ease of service. You don't wanna hafta drain the boiler to change out a reducing valve, do ya?

A valve and drain between the expansion tank and the system is also a good idea. Again, why have to drain the system when you need to change a tank, or even adjust the air charge. Google "Watts RBFF" for info on a neat 'all in one' valve made for the purpose. Legend valve makes a similar product with less features, and also a lot less $$. (The Watts is about $50 and change)

Also... you might have a hard time purging the radiant loops... you need to look at placing a 'purge station' at the proper location on the radiant loops. You have drains out there, but no 'stop valves' placed so that you can purge the loops out.

Again, maybe the drawing isn't 'literal', but you don't show a 'heat trap' on the hot water out of the indirect.

 

What I mean is, nearly all modern hydronic literature directs to pipe the boiler and system each for independent flow.
The system water should flow parallel to the boiler and not in series with it.
Another explanation would be to say the system water should flow in a large continuous circuit while the boiler should inject heat into the loop independently as needed.
That old way of having all the water flow through the boiler even while the loop temp (high limit) is satisfied is self-defeating.

Hope that is helpful and not obnoxious.

 

Look at the drawing closely. It is primary/secondary. If you follow the pipe out of the right side of the boiler, the first T you come across is the DHW loop. The next loop is the primary loop.

I think there is a problem with the DHW the way it is. Shouldn't it be off the primary loop? I think there may be a problem if the primary pump runs and the DHW pump runs at the same time during a DHW call. Unless the primary pump doesn't run when there is a DHW call.

 

Yes there are primary and secondary loops, but the boiler is in series with the primary loop...not cool.

 

No, not obnoxious... why would you think so?

I see what yer saying, but...

In this case, the boiler loop is the primary loop, and the two radiant zones are the secondary loops. I don't see a problem with that?

I think what you are suggesting would be to pipe the boiler as another secondary loop feeding the primary? Yeah, you could do that too... it would mean another pump though, and more controls.

The water heater is fine the way it is piped I think... it's just another 'zone'... and it could be prioritized by the controls if the boiler doesn't have the capacity to run the space heating and the DHW simultaneously.

 

One thing that could potentially be a problem with this setup is the 'short cycling' that was originally asked about... I would use a buffer tank too.

Master, your hydro-sep is a good idea, but I would give that H/S some steroids and morph it into a boiler buddy.

 

One more "maybe it's not a literal drawing" point...

The mixing valves are not drawn oriented properly. The mixed out of the valve is the center leg on the valve

James, while yer out buying them valves, pick up another one for the water heater outlet... or use one of them fancy Honeywell AMX300 jobbies...

 

James, I scribbled on your drawing a little bit. Up for discussion is this revision. I've included all my suggestions, and have reworked the two radiant zones into parallel connections rather than the series that you had. I honestly don't know if this is necessary, but the way it was, the second radiant zone will be presented with cooler water... the return from the first... the way I've drawn it, both zones will be working with the same temp supply water.

Remember, this is conceptual only... use at your own risk, not responsible, blah blah blah... and if anyone has any comments, feel free... I've got thick skin.

For what it's worth, I'm still somewhat concerned about short cycling with this setup.

 

Holy cow - thanks all for your input.

I too was wondering how the primary circuit could be outside of the boiler. I understand that the small loop will not hold much heat, and would rely on the boiler to fire. I see how a heat buffer tank in that loop would reduce the potential for short cycling, by only injecting additional heat into the buffer as needed. In this setup, the "primary" loop and the boiler would fire whenever either thermostat called for heat, regardless of the temperature in the primary loop. With a buffer tank, it would instead react to the temperature in the tank, independent of the heating zones? I'm beginning to see why manufacturers are pushing for these total control units, with motorized mixing valves, variable speed pumps, and temp sensors on all supplies and returns of every single loop. I just cant afford that automation right now.

I am disappointed that there are so many other things wrong with my existing system. The professional plumbers who installed this system were clearly cutting corners. The drawing is actually literal. I turned some extra elbows and tees flat and simplified the layout for clarity, but the orientation and location of all components is accurate to the existing setup.

1. There are no isolation valves (or isolation flanges) on any circulator.

2. Air tank is indeed upside-down.

3. There is no air scoop. The air trap, expansion tank, and pressure relief valve are all on a dedicated port exiting the top of the boiler, teeing off internally from the supply header. The boiler installation manual shows only the safty relief valve on this port.

4. Pressure reducing valve does fill on the return as shown without an isolation valve.

3. No heat trap or mixing valve on the DHW output. I wonder why I never noticed that before. I will check to see if it is simply storing lower temp water for anti-scald. I know I could easily extend the heat output and boost the efficiency with the mixing valve. In that case, the heat trap is to prevent heat loss by conduction through the body of the mixing valve into the cold supply? Is there another way to do this?

4. Circulators DO pump towards the boiler and are very close to the return as shown. Is there enough resistance inside the boiler itself to justify this? It is still fairly common to see circulators mounted this way, no? Ideally, I understand the inlets of the pumps should be as close to the expansion tank as possible (not pumping into the tank).

5. Yes - I drew the secondary loop mixing valves wrong. I was following an Uponor diagram. Thanks for that.


@NJ Trooper:
Thank you for taking the time to redraw my diagram. I really appreciate the pointers. I will update with revisions to the secondary loop plumbing. In the meantime, any recommendations for a buffer tank? Here are some photos of my sad setup:

 

Hi James, not all the things I pointed out are necessarily wrong. Some are... some are just me being a picky pain in the you know what! I say again, please do not accept my suggestions as something that must be done... I'm just a stranger on the internet who could be all wrong.

1. Isolation flanges on the circs can be considered a 'luxury'. They simply make life a whole lot easier when it's 3 AM on Sunday, Christmas Eve, in a blizzard... and yeah, ANYtime... because it makes a circ change a 15 min job instead of several hours draining and refilling a boiler.

2. Upside down tank could be a problem... probably millions installed that way and last a pretty long time. The potential does exist for trapped air rusting the tank prematurely.

3. IMHO, there is no excuse for not installing even the cheapest of cast iron air scoops. They are only around $20 or so... yes, the more exotic ones like the SpiroVent cost lots more... but they are also much more capable of eliminating air.

4. Circs: Yes, in a perfect world they would always pump away from the exptank connection... but it wouldn't be the end of the world if you left yours where they are. It will still heat the home (and it has been so far, right?).

5. It's just a drawing... now if you had installed them that way... well, that's another story!

Let's talk about the relief valve, mixing valve, and heat trap on the indirect a bit:

Looking at the install manual for your indirect, I notice that some of the pictures show a separate port for the T&P relief valve. However, the piping diagrams show the T&P installed on the outlet piping... BUT, and this is a big BUTT (sorry about the extra T)...

I do not believe that the way that relief valve is piped is proper at all, and it may in fact not even be both a TEMP AND PRESSURE valve. Or, it's a short stem T&P... and as such, I wouldn't trust it to do a proper job the way it's installed. The factory install diagrams clearly state that a "long element T&P valve" is to be used. That long element should extend INTO the tank, and the outlet water tee'd off the nipple that the relief valve is installed into. See any of the piping diagrams starting on page 11 in this PDF file:

http://www.htproducts.com/literature/lp-83.pdf

I'm kinda surprised that the install manual doesn't indicate that a heat trap should be used... most WH manufacturers are spec'ing them in these days. It is mostly for energy savings and with all the brou-ha-ha about that these days... you know what I mean...

A heat trap should really be used on the outlet, whether or not there is a mixing valve. The idea of course is to stop hot water migration by thermo-siphon out of the tank.

Many local codes are now requiring a mixing valve on new installations. Still, there are many that don't even know what a mixing valve is.

BUFFER TANK:

Look at Boiler Buddy - Overview (I believe they were recently bought by a major manufacturer, and that URL may redirect you)

The problem with your setup would seem to be that there isn't enough mass in the primary loop to prevent short cycling. The boiler will probably heat up very quickly... depending on the differential on your boiler's aquastat, you may find that it then fires again soon after burner shut down on high limit.

One possible solution would be to modify the aquastat to one with a wider differential... and (shudder the thought) the HeatManager might seem to be a good application here. This would seem to be a perfect case for it. Just don't believe the 'savings' they claim. This use of it would be strictly to eliminate short cycling.

So let's throw something like the BB in to the picture.

The boiler would pump into and out of two of the ports on the BB, and the radiant loops would pump in and out of the other two on a second primary loop. It would require yet ANOTHER circulator to pump this second loop.

You could set the system up with an aquastat on the BB that would maintain a minimum temp in the tank, but that would seem like a step backward... firing the boiler to keep the tank warm.

You could just put the tank in there with no control on it. Leave the boiler aquastat in control. The boiler would still fire only on a call for heat, or from the indirect.

Keeping the tank always warm has an advantage of 'instant' hot water supply to the radiant loops.

Staying cold start, there might be some time lag before heat is delivered to the heating loops...

Let me close this by saying: "I am NOT a system designer! Treat my posts as food for thought only!"

Also, I'm hoping that someone who DOES qualify as a designer will have some input here... particularly on the 'parallel' secondary loops as I revised in your drawing. There could possibly be absolutely nothing wrong with your first version having them in series.

 

Thank you for all of your help. I have incorporated your suggestions into a revised layout. Thanks also for spotting some other issues I will have to deal with.

Revised Layout (PDF)

I see how the boiler buddy is essentially a giant hydraulic separator and buffer tank in one. With four ports it simplifies the piping, and might even eliminate an extra loop circulator. It is also rather expensive.

Alternatively, I have seen piping diagrams for two-port buffer tanks, and for $50 I picked up a used electric water heater complete with aquastat, and T&P relief valve. Taking cues from this article, I have arranged it as an alternate heat source for the primary loop.

The tank circulator (P3) becomes the primary pump for the secondary circs (S1 or S2). When its temp runs low, the aquastat triggers P2 and P3 together with the boiler. Mixing occurs between unions A and B to recharge the buffer. Its aquastat would be set to a temperature below the boiler differential, and higher than the temp required for the secondary loops. I can manage the heat loss from the tank and adjust the boiler on/off cycle times by adjusting this aquastat, essentially limiting how many BTUs get stored there.

Also, during heavy heating loads in the dead of winter, I can flip a switch and take the buffer tank out of the equation.

What do you think of this setup vs. the boiler buddy?

I sized the secondary circs for the head loss through the radiant loops and various piping elements including the mixing valve, even when both are running. My question now is about filling, purging, and draining. Do you see any trouble spots in my piping?

1) Isolation valve & drain needed on buffer tank return?

2) Additional air elimination needed somewhere?

3) Is one drain on the secondary return and iso valves as shown sufficient to purge the radiant loops?

4) Are the integral flowchecks in S1 and S2 sufficient to prevent backflow when only one is running? Or, should there be additional back-flow preventers in the secondary loops?

As always, your experience and suggestions are highly valued.

(Alternate link to image version of layout below)